Project Summary Computational docking is an essential tool for analysis of biomolecular structure and function and for the discovery and development of new bioactive compounds. In particular, virtual screening is now widely used to discover new compounds to bind and inhibit targets of medicinal interest. The AutoDock Suite of programs is currently the most widely used, freely-available method for automated computational docking. This success is the result of several aspects of the work performed under the previous granting period: the continued development of the method to address problems of interest in the community, such as covalent docking and active site prediction, and the development of user-friendly interfaces and user support that ensure that the method is accessible to the widest community possible. As a result of this work, we have seen adoption and extension of the AutoDock suite by expert users, using the suite as a platform for research in computational chemistry and algorithm development, in parallel with widespread application by experimental chemists and molecular biologists who are not experts in computational chemistry. In the proposed work, we will develop AutoDock into a next generation tool for drug design and discovery. This will include extensions of the methods of AutoDock to address the expanding and varied needs of a large user community. We will develop new methods for ligand design that allow to generate novel compounds, employing specific constraints that will ensure that such compounds are synthetically accessible, and predicted to have desirable pharmacological properties. We will also continue and expand our strong commitment to user support, creating interfaces that streamline use of the AutoDock suite by the user community, and providing effective support and training.